AboutServicesMediaTalks Start a project
Dmitry Yudin
AI implementation expert

I implement AI that pays off — from strategy to a working platform

10+ years building cloud and AI platforms, leading the AI practice of one of Russia's largest cloud providers. I help companies find the problems where AI makes money: consulting and AI strategy, team training, generative AI MVPs and AI agents.

How I can help
About

Not "AI in general" — your process, your data, your economics

I have spent 10+ years in cloud technologies and artificial intelligence — launching full-cycle ML platforms, open source model catalogs and inference services, and leading an AI practice at a major cloud provider. My teams have delivered projects in manufacturing, retail and finance.

I have seen dozens of AI rollouts from the inside — both successful and failed. The difference is almost never the model: it is whether the problem had measurable impact, whether the data was ready, and whether the team could operate what was built. So I start with the process and the numbers, not the technology.

Every claim about results on this site is tied to practice: when I say "an MVP in 4–8 weeks", those are the actual timelines of my team's pilots.

What I do

  • AI strategy and economic impact assessment
  • Full-cycle ML platforms, MLOps and LLMOps
  • AI agents and multi-agent systems for business processes
  • Generative AI: RAG, assistants, document automation
  • Training — from the boardroom to engineering teams
10+
years in cloud technologies and AI
4–8 weeks
from problem statement to a working MVP — my team's actual pilot timelines
up to 25%
lower equipment maintenance costs in MLOps projects I supervised
Services

Five areas of work

AI consulting & strategy

Process audit, AI strategy with impact estimates, and a roadmap tied to your P&L — not to fashion.

Team training

AI literacy for executives, prompt engineering and LLM workflows for teams — on your company's real tasks.

Generative AI MVP

A pilot on your data in 4–8 weeks: assistants, RAG search, document processing. The result is metrics, not a demo.

AI agents

Agentic systems that run a process end-to-end: from request to document. With guardrails and a human in the loop.

AI platform engineering

Architecture, inference, MLOps / LLMOps, data safety. A platform your AI projects can live on for years.

Technology

What I work with

The stack follows the problem and your data requirements — from cloud model APIs to open source on your own infrastructure.

GigaChat YandexGPT Claude GPT · OpenAI Qwen Llama DeepSeek LangChain LangGraph vLLM PyTorch MLflow PostgreSQL · pgvector Kubernetes Airflow FastAPI
Contact

Let's find where AI pays off for you

Describe your challenge in a couple of sentences — I will reply personally and suggest a format. Your request lands directly in my Telegram.